Title:NEO Population, Velocity Bias, and Impact Risk from an ATLAS Analysis

Abstract:We estimate the total population of near-Earth objects (NEOs) in the Solar System, using an extensive, `Solar System to pixels' fake-asteroid simulation to debias detections of real NEOs by the ATLAS survey. Down to absolute magnitudes $H=25$ and 27.6 (diameters of $\sim 34$ and 10 meters, respectively, for 15% albedo), we find total populations of $(3.72 \pm 0.49) \times 10^5$ and $(1.59 \pm 0.45) \times 10^7$ NEOs, respectively. Most plausible sources of error tend toward underestimation, so the true populations are likely larger. We find the distribution of $H$ magnitudes steepens for NEOs fainter than $H \sim 22.5$, making small asteroids more common than extrapolation from brighter $H$ mags would predict. Our simulation indicates a strong bias against detecting small but dangerous asteroids that encounter Earth with high relative velocities -- i.e., asteroids in highly inclined and/or eccentric orbits. Worldwide NEO discovery statistics indicate this bias affects global NEO detection capability, to the point that an observational census of small asteroids in such orbits is probably not currently feasible. Prompt and aggressive followup of NEO candidates, combined with closer collaborations between segments of the global NEO community, can increase detection rates for these dangerous objects.